SURVEY OF RECREATIONAL FISHING IN TASMANIA. J.M. Lyle, S.R. Tracey, K.E. Stark & S. Wotherspoon

Transcription

1 27-8 SURVEY OF RECREATIONAL FISHING IN TASMANIA J.M. Lyle, S.R. Tracey, K.E. Stark & S. Wotherspoon October 29

2 Tasmanian Aquaculture and Fisheries Institute, University of Tasmania, Private Bag 49, Hobart, Tasmania Ph. (3) Fax (3) The opinions expressed in this report are those of the author(s) and are not necessarily those of the Tasmanian Aquaculture and Fisheries Institute. Tasmanian Aquaculture and Fisheries Institute, University of Tasmania 29 Copyright protects this publication. Except for purposes permitted by the Copyright Act, reproduction by whatever means is prohibited without the prior written permission of the Tasmanian Aquaculture and Fisheries Institute.

3 Recreational fishing survey survey of recreational fishing in Tasmania J.M. Lyle, S.R. Tracey, K.E. Stark and S. Wotherspoon Executive summary This represents the second comprehensive assessment of recreational fishing undertaken in Tasmania and builds on the National Recreational Fishing Survey (NRFS) conducted in 2-1. The same methodology developed for the NRFS was applied in the current survey but with several improvements, particularly in relation to data analysis. However, in the absence of a repeat of the national survey, the current survey was limited to fishing in Tasmania by Tasmanian residents. Although not measured, fishing by interstate visitors in Tasmania was not likely to have been significant given that non-resident fishers accounted for just 3% of the total effort (fisher days) for Tasmania during 2-1. Information about participation rates and the demographic profile of recreational fishers was derived from a general population telephone survey involving over 34 Tasmanian households. This was followed by a telephone-diary survey involving over 1 households (almost 3 persons) for which fishing activity was monitored over a 12 month period. Response rates across all facets of the study were exceptionally high, giving considerable confidence to the data quality. In the 12 months prior to November 27 it was estimated that over 118, Tasmanian residents aged five years or older fished at least once, representing an overall participation rate of 26%. By region, residents of the Southern statistical division had the highest participation rate at 33%, which compared with 24-27% for the other Tasmanian regions. Recreational fishing was more popular among males, with a state-wide participation rate of 35%, compared with about 18% for females. Participation rates varied with age, 5-14 year olds having the highest rate of participation (38%) although the greatest numbers of fishers were in the 3-44 years age group. Participation rates generally declined with increasing age, but especially in the 45 years and older age groups. As this survey was designed to provide a big-picture perspective of the recreational fishery, it is important to recognise that comparatively rare or highly specialised activities, which within the context of the overall recreational fishery are minor components, may not be well represented. In such instances estimates of catch and effort tend to be imprecise and alternative, targeted surveys would be required to provide a more reliable assessment of such activities. For the above reasons, aggregation of some regions and species has been necessary when reporting findings. Information about recreational fishing catch and effort was monitored between December 27 and November 28, inclusive. Almost 128, Tasmanian residents were estimated to have actually fished in Tasmania during this period, slightly more TAFI Report Page i

4 Recreational fishing survey 27-8 than during the previous 12 months. These fishers accounted for about.64 million fisher days of effort. The median number of days fished in Tasmania by Tasmanians was five days per fisher, though the distribution of effort was highly skewed, with just 2% of fishers contributing 56% of the total effort. Overall, one in four fishers fished at least once in freshwater while the vast majority (88%) fished at least once in saltwater. About one quarter of the total effort occurred in freshwater, saltwater fishing (including estuaries) accounted for the remainder. Freshwater fishing in lakes and dams accounted for about three times the level of effort in rivers while the majority of the saltwater fishing occurred in inshore coastal waters, with estuarine fishing of secondary importance. Fishing in offshore waters (>5 km off the coast) was a comparatively minor activity. Line fishing was the dominant activity undertaken, pursued on 87% of all days fished: that is almost.56 million fisher days, representing 1.8 million hours of effort. This was followed by pot fishing (8%), dive harvesting (5%) and the use of gillnets (3%). A range of other fishing methods were also reported, including the use of spears, seine or bait nets, and hand collection, but these activities were of minor significance by comparison. A wide variety of fish species was caught by recreational fishers during 27-8, with a total of 1.62 million finfish (excluding small baitfish) retained and 1.24 million finfish released or discarded. Flathead (mainly sand flathead) represented almost two-thirds of the total finfish catch numbers, with an estimated 1.7 million kept and.74 million released. Other finfish species or species groups of significance included trout (157, kept and 15, released), Australian salmon (11, kept and 78, released), gurnard (13, kept and 67, released), and black bream (13, kept and 35, released). Overall, 43% of all finfish captured were released or discarded; with low rates of release (<1%) for species such as blue warehou and flounder; intermediate rates (1-3%) for garfish, trumpeters, Atlantic salmon and jack mackerel; moderate rates (31-5%) for flathead, trout, Australian salmon, tuna, mullet, barracouta, silver trevally, jackass morwong, eels, river blackfish and redfin; and high rates (>5%) for black bream, wrasse, gurnard, sharks and rays, whiting, cod, pike and leatherjackets. Reasons for release were varied, with size (under legal size or too small) being an important factor for species such as flathead, Australian salmon, silver trevally, mullet and jackass morwong; poor eating qualities were identified as an important factor for release of barracouta, redfin, cod, wrasse, leatherjackets and gurnard; while catch and release (sport) fishing was an important factor for the release of black bream, trout and tuna. Sharks and rays tended to be released or discarded because of poor eating qualities (e.g. dogfish, draughtboard shark) and/or due to regulation (prohibition on retaining sharks from shark refuge areas). Recreational fishers also caught a variety of shellfish and other invertebrate species. Comparatively high catches of squid, namely Gould s squid (73, kept) and southern calamari (4, kept), were taken along with rock lobster (72, kept), abalone (64, kept) and scallops (397, kept). Amongst these taxa, rates of release were low for the squids, scallops and abalone, and moderate for rock lobster. TAFI Report Page ii

5 Recreational fishing survey 27-8 There was a high level of fishery specialisation for species such as tuna, trout, flounder, rock lobster and scallops; these species were taken almost exclusively by targeted effort rather than incidental capture. Other species that tended to be caught primarily as a result of targeted effort included black bream, flathead, garfish and abalone, also implying a level of fishery specialisation for these species. By contrast, jackass morwong, jack mackerel, leatherjackets, wrasse, gurnard, cod and eels were virtually never targeted, which for several of these species is consistent with the fact that they were held in low esteem by fishers. By applying average weights it was possible to approximate harvest weights and compare recreational and commercial fisheries production. The annual recreational harvest of flathead was estimated at 292 tonnes, four times greater than the commercial catch of flathead taken from state fishing waters. By weight, other species of importance included tuna (145 tonnes), Australian salmon (48 tonnes), southern calamari (45 tonnes), Gould s squid (37 tonnes) and the trumpeters (19 tonnes). As a contributor to total harvest, the share taken by the recreational sector was similar or larger than that taken by the Tasmanian commercial scalefish fishery for flathead, flounder, mullet, cod, barracouta, silver trevally, jackass morwong, and Gould s squid. Conversely, the recreational harvest represented a minor component (<15%) of the total catch for species such as Australian salmon, whiting, garfish, wrasse and jack mackerel. Catch composition was influenced by many factors, including the water body fished and the fishing method. Trout dominated finfish catches (kept and released numbers) in freshwater (>8%), with redfin, Atlantic salmon and blackfish of secondary importance in the lake and dam fisheries, and redfin and blackfish in the river fisheries. Flathead and Australian salmon dominated estuarine and inshore coastal catches (collectively >75%), with black bream and flounder of secondary importance in the estuarine fishery, and gurnard and wrasse in the inshore coastal fishery. Tuna, flathead and gurnard (mostly ocean perch) were the main species taken in the offshore fishery. The finfish catch taken by line fishing was dominated by flathead (66% of total numbers), followed by trout (1%), Australian salmon (7%), gurnard (3%) and bream (2%). By contrast, trumpeter (mainly bastard trumpeter) (27%), blue warehou (1%), sharks and rays (9%), mullet (9%) and Atlantic salmon (7%) were the main species caught by gillnets, and flounder were mainly taken by spear. Gould s squid and southern calamari were mostly taken by line methods whereas rock lobster were caught using pots, dive collection and rings, with abalone and scallops more or less exclusively harvested by dive collection. The east and south-east coasts of Tasmania were a particularly significant for flathead, black bream, tuna, Gould s squid, southern calamari, rock lobster and abalone, with the south east especially important for flounder. By contrast, Australian salmon and mullet catches were concentrated off northern Tasmania. The inland trout fishery was focused largely in the Central Plateau lakes, especially Arthurs Lake and Great Lake, with catches from rivers of secondary importance. Trout catches from the other inland regions were similar in magnitude. Seasonally, catches of flathead, trout, Australian salmon, black bream, Gould s squid, southern calamari, rock lobster and abalone peaked during summer and autumn. Tuna were restricted to summer-autumn, with a strong peak in February-March. Flounder TAFI Report Page iii

6 Recreational fishing survey 27-8 catches were highest in late autumn. Catches of each of the major species tended to be low during winter and early spring, reflecting the generally lower levels of fishing activity during that period. The saltwater fisheries off western and northern Tasmania involved a significant shorebased component whereas shore-based fishing was less important compared with boatbased fishing off eastern and south-eastern Tasmania. Comparison with 2-1 To facilitate valid comparisons between 2-1 and 27-8, NRFS data were reanalysed using the analytical approach developed for the current study. In terms of participation, the number of recreational fishers in Tasmania has remained relatively constant since 2. However, when population growth is taken into account, the actual participation rate experienced a significant decline, from just over 29% in 2 to 26% in 27. This decline was experienced in all regions of the state and was more pronounced amongst males than females. Participation rates were also lower across all age groups, with the exception of the 6 years-plus age group. When broader demographic trends are taken into account, specifically dominance of the baby-boomer generation (mostly year olds in 27) and the sharp decline in participation rates amongst the 45 years-plus age groups, our results suggest that overall participation in recreational fishing will continue to decline unless there is growth, or at least maintenance, of involvement in fishing amongst the younger age groups. Overall effort (fisher days) was 14% lower in 27-8 compared with 2-1. This decline was exclusively linked to a reduction in shore-based fishing effort during The most marked declines were experienced in the Eastern inland, West North coast and Derwent regions, mainly due to lower levels of shore-based fishing activity in Reflecting the decline in effort there was also a reduction in overall catch (kept and released) numbers for finfish in 27-8, to about 81% of the equivalent estimate for 2-1. While the catch composition and relative importance of the key species was generally consistent between surveys, there was variability in catch levels for many species; linked in part to differences in effort, changes in fishing practices and species availability. Flathead catches were very stable (within 5%) between the two surveys while there was a moderate increase (17%) in the estimated number of trout caught in By contrast, the 27-8 catch of Australian salmon was less than half that estimated for 2-1. Catch increases were experienced for tuna and Gould s squid, both linked to greater availability during Minor increases were also evident for southern calamari and whiting, the former apparently linked to increased popularity of the species amongst recreational fishers. Release rates were generally higher in 27-8, which for species such as trout, black bream and tuna was consistent with a trend towards increased catch and release (sport rather than consumptive) fishing. For other species, this change may reflect improved adherence to size limits and/or the impact of decreased bag limits that have been implemented since 21, as well as response to education programs aimed at encouraging fishers to take only what they need for a feed. TAFI Report Page iv

7 Recreational fishing survey 27-8 In summary, this study has highlighted the complex and dynamic nature of the recreational fishery and emphasises the need for managers, stakeholders and fisheries scientists to consider management, research and planning issues at appropriate regional and temporal scales. This survey represents a significant step towards achieving this goal, providing an important baseline against which future developments and trends in the fishery can be evaluated. TAFI Report Page v

8 Recreational fishing survey 27-8 Table of Contents EXECUTIVE SUMMARY... I 1. INTRODUCTION MATERIALS AND METHODS SURVEY SCOPE SURVEY METHODOLOGY Screening survey Diary survey Non-intending fisher call-backs Wash-up survey On-site survey DATA ANALYSIS Data expansion Statistical uncertainty REGIONS Sampling strata Fishing regions FISHING EFFORT FISHING METHODS CATCH SAMPLE AND RESPONSE PROFILES SCREENING SURVEY DIARY SURVEY NON-INTENDING FISHERY CALL-BACKS FISHER CHARACTERISTICS PARTICIPATION RATES AGE AND GENDER FISHING EFFORT DAYS FISHED WATER BODY FISHING METHOD FISHING REGION FISHING PLATFORM CATCH TOTAL CATCH, HARVEST AND RELEASE/DISCARDS Reasons for release Targeted fishing Harvest weights CATCH BY WATER BODY CATCH BY METHOD Line fishing CATCH BY PLATFORM KEY SPECIES FLATHEAD TROUT AUSTRALIAN SALMON MULLET FLOUNDER BLACK BREAM TAFI Report Page vi

9 Recreational fishing survey TUNA GOULD S SQUID SOUTHERN CALAMARI ROCK LOBSTER ABALONE REGIONAL FISHERIES INLAND FISHERY WEST COAST WEST NORTH COAST TAMAR ESTUARY EAST NORTH COAST NORTH EAST COAST CENTRAL EAST COAST SOUTH EAST COAST NORFOLK-FREDERICK HENRY BAY DERWENT ESTUARY D ENTRECASTEAUX CHANNEL COMPARISONS WITH FISHER CHARACTERISTICS Participation rates Age and gender Participation rates since FISHING EFFORT CATCH SUMMARY AND CONCLUSIONS ACKNOWLEDGEMENTS... 7 REFERENCES APPENDICES TAFI Report Page vii

10

11 Recreational fishing survey Introduction With growing awareness of the significance of recreational fishing in the early 199s, a national policy for recreational fishing was developed in Australia. The policy was released in 1994 and endorsed the principle that fisheries management decisions should be based on sound information including fish biology, fishing activity, catches and economic and social values of recreational fishing (NRFWG, 1994). The policy recommended that a national survey of recreational fishing be undertaken once every five years. Recognition was also given to public concern over the then poor quality of data on recreational fishing. Following extensive consultation and development phases, the Commonwealth, state and territory fisheries agencies implemented the National Recreational Fishing Survey (NRFS) in 2. The principal objectives of the NRFS were to determine participation rates in recreational fishing; profile the demographic characteristics of recreational fishers; quantify recreational catch and effort; collect data on expenditure by the recreational fishing sector; and establish attitudes and awareness of recreational fishers to issues relevant to the fishery (Henry and Lyle, 23). The NRFS was implemented as a series of state-wide surveys using a common methodology, having the advantage of providing comparable information Australia-wide as well as including the activity of visiting fishers. In addition to nationally aggregated information, Henry and Lyle (23) provided summary statistics for each of the states and territories. In Tasmania it was established that the rate of recreational fishing participation was higher than the national average (29.5% compared with 19.5%), the average number of days fished per fishers was slightly higher than the national average (6.55 compared with 6.1 days per year) while the average annual expenditure per fisher 1 was lower than the national average ($416 compared with $552). Lyle (25) provided a more in-depth analysis of the NRFS dataset as it pertained to Tasmania, examining the data at a regional level, for key species and by fishing methods. As such this provided the first comprehensive state-wide assessment of recreational fishing and represents an important benchmark against which future trends in participation, catch and effort can be measured. Prior to the NRFS, only limited information was available about recreational fishing in Tasmania. In 1983 an Australian Bureau of Statistics (ABS) household survey provided some general statistics on recreational fishing but no estimates of catch or effort (ABS, 1984). At that time about one third of all persons aged 15 or older were engaged in some form of recreational fishing activity. A subsequent survey of home food production estimated home seafood production for the year ending April 1992 at over 1 tonnes for finfish (including trout), 6 tonnes for rock lobster and 25 tonnes for abalone (ABS, 1994). There have been several other attempts to gather information about the recreational fishery in Tasmania. A survey of recreational fishing in the Derwent Estuary conducted 1 Based on attributed expenditure on selected recreational fishing-related items (refer Henry and Lyle, 23). TAFI Report Page 1

12 Recreational fishing survey 27-8 in 1984 provided some information about motivation, expenditure and catch rates, but gave no information on effort levels, total catch or catch composition (Winter,1985). Lyle (2) conducted a comprehensive survey of licensed marine recreational fishing that provided catch and effort information for the recreational net, rock lobster and abalone fisheries between 1996 and There have also been several other surveys of the recreational rock lobster and abalone fisheries (Forward and Lyle, 22; Lyle et al., 25; Lyle and Morton, 24, 26; Lyle, 28) and limited surveys of game fish catch and effort (Smith, 1994; Evans, 1995; Morton and Lyle, 23). In terms of the inland trout fishery, catch and effort trends have been monitored over several years using mail survey methods (Davies, 1995). In the absence of plans to repeat the national survey, the present study was developed to provide up-to-date big-picture information on recreational fishing in Tasmania. This information includes state-wide participation rate and demographic profile of recreational fishers, and catch and effort estimates for key methods, regions and species. However, in providing this big-picture perspective, it is also recognised that more targeted surveys will be required to provide greater accuracy and precision for specialised or localised activities such as rock lobster fishing, game fishing, gillnetting, etc. By adopting essentially the same survey methodology developed for the national survey, valid comparisons can be made with information collected in 2-1. Several improvements to the statistical analyses were, however, implemented in the present study and therefore data from 2-1 have been re-analysed accordingly. In conjunction with the previous survey, it is intended that the current study will represent the start of a series of periodic state-wide fishing surveys that will be used to monitor major developments, trends and general status of recreational fishing in Tasmania into the future. TAFI Report Page 2

13 Recreational fishing survey Materials and methods The primary data collection was based on a telephone-diary approach, an off-site methodology developed to provide cost-effective data over large spatial scales, such as for the entire state. A detailed description of the telephone-diary design philosophy and methodology is provided in Lyle et al. (22a) and Henry and Lyle (23). Data analysis procedures are described in detail by Lyle et al. (29) and have been undertaken using the statistical computing language R (R Development Core Team, 28). An overview of the survey methodology and data analysis is provided below. 2.1 Survey scope The survey encompassed the private dwelling resident population of Tasmania, aged five years and older, and their recreational fishing activity. In this context recreational fishing was defined broadly as the capture or attempted capture of aquatic animals in Tasmanian waters (freshwater, estuarine, marine) other than for commercial purposes. All recreational fishing techniques and harvesting activities, including dive and hand collection, the use of pots, nets and spears in addition to line fishing, were considered inscope. In contrast to the 2-1 survey, fishing activities by non-tasmanian residents in Tasmania and fishing by Tasmanians in other states of Australia was considered out-ofscope. 2.2 Survey methodology The telephone-diary methodology involved a two-phase survey design, the principal components being an initial screening phase to gather profiling information from a sample of the population and a subsequent, intensive phase, in which respondents provided detailed catch and effort information over a period of time. In this second phase, respondents were encouraged to use a simple diary to record key fishing data and were contacted regularly by survey interviewers, who were responsible for collecting this information. The underlying design philosophy is focussed on minimising respondent burden and maximising response and data quality. Additional survey components included non-intending fisher call-backs, a fisher motivation, awareness and attitude survey, and an on-site survey. The non-intending fisher call-backs involved a sample of households that had indicated at screening that none of the residents were likely to do any recreational fishing during the diary period. This component was designed to identify and account for unexpected fishing that may have occurred during the diary period. Motivation, awareness and attitudes to fishingrelated matters was assessed for diary participants at the end of the diary period in a wash-up survey. On-site (creel) surveys were also conducted to assess fish identification skills of recreational fishers, determine the size distribution of common species and provide independent verification of certain recreational fishing activities. TAFI Report Page 3

14 Recreational fishing survey Screening survey The primary role of the screening interview was to collect profiling information for all household members as well as establishing eligibility to participate in the follow-up diary phase. Profiling information was important not only to characterise the sample population but also to examine issues relating to representation and response. The screening survey was administered as a structured interview by telephone on a random sample of Tasmanian households. The white pages directory provided the sample frame, with obvious business numbers, non-private dwellings and multiple listings removed. For each telephone number the suburb was also noted enabling the selection to be assigned to a Local Government Area (LGA) and Statistical Division (SD). Stratified random sampling was undertaken with a higher sampling rate for the Southern SD, lower but equal sampling rates for the Northern and Mersey-Lyell SDs and the lowest rate for the Greater Hobart SD. Within each SD, care was taken to ensure that the proportional breakdown of the sample at the LGA level did not differ significantly to the known proportion of private dwellings based on Australian Bureau of Statistics (ABS) data. In addition to landline numbers, 14% of selected listings were represented by mobile-only numbers. In order to minimise non-contacts, at least 1 calls were made to each live telephone number. Disconnected numbers, business and facsimile numbers were treated as sample loss and not replaced. The screening survey was conducted during October and November 27. Within each responding household, the demographic profile (age group and gender) of all usual residents, involvement in recreational fishing over the previous 12 months and likelihood (expectation) of doing any recreational fishing in the following 12-months for residents aged five years or older was established. All respondents who had fished during the 12 months prior to interview were asked whether they had fished in fresh and/or saltwater, whether they had fished interstate and to estimate how many days they had fished in the previous 12 months. This latter detail was used as an index of avidity rather than a direct or accurate measure of prior fishing activity, with fishers broadly classified as infrequent, occasional and regular based on the number of days reported. Previous and intending fishers were also asked whether they were members of fishing clubs or associations; ethnicity was established based on languages other than English spoken at home. Boat ownership was also established for all households, regardless of whether they were fishers or not. All households in which at least one member (regardless of prior fishing history) expressed a likelihood of going fishing during the following 12 months were considered eligible for the second (diary) phase of the study Diary survey All households identified as eligible for the diary survey were invited to participate in this phase of the study. Fishing activity of household members aged five years and older was monitored between December 27 and November 28, inclusive. TAFI Report Page 4

15 Recreational fishing survey 27-8 The approach taken in this survey differed to conventional angler diary surveys in two important ways: first the diary was employed more as a memory jogger than a logbook; and second, responsibility for data collection rested with the survey interviewers and not the diarists. Typically, diary survey response rates are low and data quality can suffer in terms of completeness, generality and consistency. Since the burden of maintaining the diary rests with the respondent, instructions may be misinterpreted and data may be incomplete or ambiguous. The need to periodically remind respondents to submit documentation creates a further problem, whereby information that has not been diarised must be collected on the basis of recall, if at all. By contrast, the telephone-diary approach, a form of panel survey, effectively transferred the burden of data collection from the respondent to the survey interviewer. Data collection was undertaken by brief telephone interview in which trained interviewers recorded details of any fishing that had occurred since the last contact. The level of fishing activity determined the frequency of such contact but, as a general rule, respondents were called at least once a month even if no fishing was planned. After receiving the diary kit which included the diary, a colour species identification guide to the common species, and a survey cover letter, data requirements were explained to respondents in a brief interview and then the next contact arranged. Respondents were encouraged to record basic information in their diaries, such as date, location, start and finish times, and catch and release numbers. More detailed data, such as target species, fishing method, platform (boat or shore), water body type (river, lake, estuary, coastal, offshore, etc), and reason(s) for release, for each individual fishing event were collected and recorded during the telephone interview. By maintaining regular contact, usually within a couple of weeks of any fishing activity, details of any non-diarised fishing were obtained with minimal concern in relation to recall bias. Furthermore, interviewers were able to immediately clarify ambiguities and ensure completeness of information. This in turn, provided for greater data utility, for example fishing effort could be apportioned between target fisheries, methods, fishing platform, and so on Non-intending fisher call-backs The objective of the call-backs was to account for those persons who may have unexpectedly dropped-in to the fishery, providing symmetry for those persons who unexpectedly dropped-out of the fishery; the latter group being identified as diarists who, despite indicating an expectation to fish, did no fishing during the diary period. A random sample of households which at screening had indicated no intention to go fishing during the diary period (i.e. not eligible for the diary survey), was re-contacted shortly after the diary period in early 29. Whether any fishing had occurred during the diary period was established in a brief telephone interview, with particular care to identify whether there had been a change in household (e.g. telephone number re-allocated) and that household members were the same as those at screening. Further details were collected from those households in which fishing was reported, including demographic profile (age group and gender), whether individual members had fished in Tasmania and/or interstate, in salt and/or freshwater, number of days fished during the 12 months of the diary period and whether key species were caught and kept. Respondents who were TAFI Report Page 5

16 Recreational fishing survey 27-8 identified as not being residents of the household at the time of screening were excluded from the analysis Wash-up survey The survey was conducted with diarists at the end of the diary period and was designed to assess a range of information, including fisher motivation and attitudes to and awareness of fisheries issues. All such information was obtained from the main/key fisher in the household, aged 18 years or older. The survey also confirmed with respondents the completeness of the data for each household member, whether they had reported fishing or not. The results of this survey will be provided in future reports On-site survey A small team of survey staff conducted interviews with fishers at several south-east and east coast boat ramps between December 27 and April 28. In addition to collecting information about the fishing trip (fishing location, methods, time fished and catch), the ability of recreational fishers to correctly identify their catch to the taxonomic level required for data analysis was assessed by comparing the species reported by the fishers with the actual identity of the catch based on observation. Where feasible size composition information was also collected and coupled with other data sources, this was used to estimate the mean size and weight for key species. In combination with harvest estimates (numbers), mean weights can be applied to estimate recreational catches by weight for comparison with commercial catch statistics. 2.3 Data analysis Data expansion Data analysis was based on a stratified random survey design using single stage cluster sampling, with the household representing the primary sampling unit (PSU) and residents within the household the secondary sampling unit (SSU). In determining household and individual expansion factors, an integrated approach was applied that adjusted for nonresponse and calibrated against population benchmarks (Lyle et al., 29). Adjustment for non-response at screening was based on fishing propensity determined amongst households that refused to complete the screening interview, but at least answered the question about whether or not household members had fished in the previous 12 months. Calibration against ABS estimated resident population (ERP) data for Tasmania as at July 27 was implemented taking account to household size and demographics. Using diary phase uptake and completion rates for eligible households, further non-response adjustment was applied to expansion factors in calculating catch and effort information. This adjustment was made sensitive to the avidity classification for the household (the maximum avidity index for a member of the household determined at screening) and region of residence (stratum). TAFI Report Page 6

17 Recreational fishing survey 27-8 Not all eligible fishers actually fished during the diary period and these in effect represented unexpected drop-outs from the fishery. In order to take into account unexpected drop-ins to the fishery, a final adjustment was necessary and was based on the non-intending fisher call-back survey. This adjustment was made sensitive to the avidity index reported for drop-ins and region of residence (stratum). A full account of the analytical process is provided by Lyle et al. (29). Unless otherwise indicated, parameter estimates provided in this report are based on expanded data, scaled-up to represent the population rather than the sample from which they were derived Statistical uncertainty As a consequence of surveying a sub-sample rather than the entire population of fishers, all parameter estimates have some statistical uncertainty; that is, estimates may differ from those that would have been produced had the entire population been surveyed. This uncertainty can be expressed in terms of standard error (SE), which indicates the extent to which the estimate might have varied from the true population value due to chance because only a sample was included in the survey. There are about two chances in three (67%) that sample estimates will vary by less than one SE and about 19 chances in 2 (95%) that the difference from the true population value will be less than two SEs. It should be noted that as survey data are disaggregated, for example by region or method, SEs expressed as a percentage of the estimate (known as relative standard error or RSE) will increase and there may become a point where the disaggregated estimates become unreliable. In interpreting survey estimates consideration needs to be given to: a) the magnitude of the RSE and b) the actual number of households that contributed records to the estimate. Estimates with RSEs of 4% or greater (implying a 95% confidence range of ± 8%) have been highlighted and are regarded as imprecise. Estimates derived from records involving fewer than 3 households have also been highlighted since they may be particularly influenced by the activities of very few fishers and hence may not be representative. 2.4 Regions Sampling strata Initial household selection (i.e. telephone number) was based on stratified random sample design using the four ABS Statistical Divisions (SD) as strata: Greater Hobart, Southern, Northern, and Mersey-Lyell (Fig. 1). In describing household and population characteristics data have been analysed at stratum (SD) and state levels. TAFI Report Page 7

18 Recreational fishing survey 27-8 Mersey- Lyell Mersey- Lyell Northern Northern Southern Hobart Fig. 1 Map of Tasmania showing ABS Statistical Divisions used for sample stratification Fishing regions During the diary survey, interviewers classified the location of each fishing activity (event) into one of 23 fishing regions as employed in the NRFS. However, unlike the NRFS, the reported fishing location (text) was routinely recorded in the database, both as a validation tool and to provide added flexibility in ongoing analysis work. For reporting purposes it has been necessary to collapse some regions to ensure that a minimum of 25 fishing events (i.e. raw unexpanded data) occurred in each reporting region. The fishing regions used for data reporting include inland, selected estuarine, and coastal regions as indicated in Fig. 2. Other fishing location information was also collected in the diary survey in terms of water-body type: marine waters > or < 5kms from the coastline; estuarine waters; freshwater rivers; and freshwater lakes/dams, public or private. TAFI Report Page 8

19 Recreational fishing survey 27-8 W North coast E North coast West coast Tamar Western Eastern Central N East coast Central East coast DEC NFHB Derwent S East coast Fig. 2 Map of Tasmania showing analysis regions used for reporting fishing activities. Fishing regions - NFHB Norfolk and Frederick Henry bays; DEC D Entrecasteaux Channel. 2.5 Fishing effort Fishing information was collected on an event basis, where an event was defined as a discrete fishing episode and the actual household member(s) involved in the event recorded. Separate fishing events were defined where there was a change in fishing region or water body type, target species and/or fishing method. In this way a day s fishing trip could comprise more than one event; for instance, fishers may gather bait prior to fishing for flathead. Both the gathering of bait and the subsequent fishing were considered to be separate events since the effort expended in the capture of bait cannot be attributed to the capture of flathead and vice versa. Similarly, the use of passive fishing gear, such as rock lobster pots or gillnets, at the same time as line fishing were recorded as separate fishing events. The delineation of fishing activity in this manner provided an ability to analyse effort (and catch) on the basis of fishing method and target species/fishery. Furthermore, three measures of effort could be defined, namely fisher days (i.e. separate days on which some form of fishing was undertaken by a fisher), fishing events, and hours fished. It should be noted that person-based effort has been calculated for this report. For active fishing methods such as line fishing and dive harvesting this is clearly appropriate, but where shared or joint activities occurred, such as fishing with lobster pots or using graball nets, this can over estimate effort. In such instances, effort was calculated as the number of pots/nets used divided by the number of persons who participated in the fishing activity on a given day, providing an effort measure of the number of person pot/net days of effort. TAFI Report Page 9

20 Recreational fishing survey Fishing methods A variety of fishing/harvesting methods were identified by diarists but for the purposes of analysis the following reporting categories have been defined: line fishing (bait and/or lure/jig/fly or set lines); rock lobster pot; gillnet (graball and mullet nets); dive collection (includes underwater spearfishing and hand collection on snorkel, scuba or hookah), and other methods (including surface hand spearing, lobster rings, beach seine, dip nets, cast nets, bait nets, hand collection and the use of spades). Grouping of the minor fishing methods in this way ensured that overall there was a minimum of 2 events in each method reporting category. 2.7 Catch A Species Identification Guide including clear colour images was provided to all diarists to optimise the accuracy of species identification in the survey. A key factor here is that the resolution required for individual species must recognise the identification capabilities of fishers, on a lowest-common-denominator basis. Although excellent reporting precision can be achieved at the species level in some instances (confirmed through onsite surveys - Lyle and Campbell, 1999; Lyle et al., 22b), species groupings were required where fishers could not reasonably be expected to delineate particular species, even with the aid of the identification booklets. For example, icon species such as striped trumpeter were readily recognisable whereas identification to species level for flounder was less certain, even though flounders could be readily distinguished from other groups of fish. For the purpose of reporting catches, species (e.g. Australian salmon, black bream) or taxonomic groupings (e.g. flathead, trout, flounder) have been used in most instances. Nonetheless, several species or species groups were represented by very few records and thus it was necessary to pool these into broader taxonomic categories for analysis (e.g. sharks and rays, other scalefish). A listing of taxa reported in catches and the catch analysis groupings are provided in Appendix 1. Catches were reported as numbers of individuals kept or harvested and numbers released or discarded by species. In a small number of instances, respondents reported catches of small and generally abundant species (e.g. whitebait) in units of weight or volume. These catches were later converted to numbers using agreed number to weight or volume conversions 2. 2 These conversion factors were based on best guess estimates. TAFI Report Page 1

21 Recreational fishing survey Sample and response profiles 3.1 Screening survey Table 1 provides a summary of the number of private-dwelling households in Tasmania as at July 27 (based on ABS ERP data), sampling details and the response profile relating to the screening survey. Since sampling was undertaken without replacement for sample loss (e.g. disconnected numbers, non-private dwellings including businesses, nursing homes, etc), the net sample was reduced from a gross sample of 4632 to 482, of which 3451 households (84.5%) fully responded to the screening survey. Response rates were relatively consistent across all sampling strata. Overall, demographic profiling information was collected from 833 persons aged five years or older. Non-response was due to refusals (8.6% overall), non-contacts (6.4%) and other non-response (.6%), such as language or communication difficulties. The refusal group was equally split between partial refusals (where at least the substantive question relating to previous household fishing was answered) and full refusals where no information was provided. Table 1 Tasmanian private dwelling population (number of households), survey sample size, and responses to the screening survey by stratum Net sample- initial sample less sample loss Statistical division Households Initial sample Net sample Response Refusals Noncontact Other nonresponse % response Greater Hobart 82, Southern 14, Northern 55, Mersey-Lyell 44, Total 197, Diary survey Table 2 summarises response details as they related to the diary survey. Of those households identified at screening as having at least one resident with an intention to do some recreational fishing during the diary period (December 27 to November 28), almost 85% fully responded to the diary survey. In total, 137 Tasmanian households, representing 2936 persons aged five years and older, participated in the diary survey, with consistent response rates across all strata. The diary survey yielded a total of 1,148 fishing events. Based on those households which initially agreed to take part in the diary survey (182), the effective diary completion rate was 95.8%. A similar diary completion rate (96.8%) was achieved for the Tasmanian component of the NRFS (Lyle, 25). TAFI Report Page 11

22 Recreational fishing survey 27-8 Overall, by comparison with other general population surveys and traditional mail-back diary studies, the response rates achieved in all components of this study are exceptional and represent an important indicator in terms of the efficacy of the survey instrument. Table 2 Diary survey response profile by stratum Statistical division Eligible households Completed diary % response Greater Hobart Southern Northern Mersey- Lyell Total Non-intending fishery call-backs Response rates for the non-intending fisher call-backs are presented in Table 3. Approximately half of the 2225 households that indicated no intention to go fishing during the diary period were selected at random to be followed up at the end of the diary period to ascertain whether any unexpected fishing had occurred. When sample loss (disconnected numbers, different household) was taken into account, an overall response rate of 92% was achieved for this segment of the study, again with consistently high response rates between strata. Table 3 Response profile (household) to the non-intending fisher call-back survey by stratum Other Statistical Initial Net Noncontact response % Response Refusals nonresponse division sample sample Greater Hobart Southern Northern Mersey- Lyell Total 1,99 1, TAFI Report Page 12

23 Participation rate (%) Number of fishers Recreational fishing survey Fisher characteristics The following analyses are based on information derived from the screening survey of Tasmanian residents and are expanded, with non-response adjustments, to represent the resident private dwelling population of Tasmania. 4.1 Participation rates An estimated 118,399 (SE 3363) Tasmanian residents aged five years or older fished at least once in the 12 months prior to November 27 (Appendix 2). About 43% of the fishers resided in the Greater Hobart, 27% in the Northern, 2% in the Mersey-Lyell and 9% in the Southern SDs (Fig. 3A). The effective participation rate (proportion of resident population) was 26.1% (SE.7%) of Tasmanians; regional participation rates were relatively consistent between 24 and 27% apart from the Southern SD, where 33% of the population engaged in some form of recreational fishing (Fig. 3B). A 6 5 B Greater Hobart Southern Northern Mersey-Lyell Greater Hobart Southern Northern Mersey-Lyell Fig. 3 Fishing participation in the 12 months prior to November 27 by SD for persons aged five years or older: A) Number of persons; and B) proportion of the resident population. Error bars represent one standard error and the dotted line represents the participation rate for Tasmania as a whole. TAFI Report Page 13

24 Participation rate (%) No. fishers Recreational fishing survey Age and gender Recreational fishing was more popular among males, with 34.9% (SE.9%) of the male and 17.5% (SE.8%) of the female resident population in Tasmania aged five years or older participating in recreational fishing in the 12 months prior to November 27 (Appendix 3). By numbers, almost twice as many males (78,177; SE 212) than females (4,222; SE 177) did some form of recreational fishing. The predominance of males involved in fishing, by number and proportion of population, was evident across all age groups (Fig. 4) and by statistical division (Appendix 3). The number of persons (both males and females) who fished generally increased with age up until 3-44 years after which numbers declined quite markedly, such that by 6 years or older comparatively few persons were involved in fishing (Fig 4A, Appendix 3). By contrast, participation rates generally declined with age, falling to just 12.8% amongst persons aged 6 years plus (21.5% of males and 4.8% of females) (Fig 4B, Appendix 3). Thus, although the greatest numbers of fishers were in the 3-44 years age group (19,545 males and 1,785 females), the highest participation rates occurred in the 5-14 years age group, 37.7% overall (44.9% for males and 3.1% for females). A 25 2 Male Female B to to 29 3 to to 59 >6 Age group Male Female 5 to to 29 3 to to 59 >6 Age group Fig. 4 Fishing participation in the 12 months prior to November 27 by age group and gender by Tasmanian residents aged five years or older: A) number of persons; and B) proportion (%) of the resident population. TAFI Report Page 14

25 Recreational fishing survey Fishing effort Fishing effort is used to describe the pressure applied to a resource by fishers and to derive (with catch data) indices of resource abundance and fishing success. The response of fish populations to variations in fishing effort represents an important foundation for stock assessment. For the purposes of this report only effort undertaken by Tasmanian residents in Tasmania and its adjacent waters has been considered. Effort can be described in several ways, for instance on the basis of fisher days (regardless of time fished on the day), hours fished or events (as defined in this study). For this report, the primary effort metric used is fisher days, noting that a fisher day of effort can be disaggregated by fishing region, water body type, platform and/or method. Based on reported activity by diarists, with appropriate non-response and drop-out and drop-in adjustments, it was estimated that overall 127,781 Tasmanian s fished between December 27 and November 28 (Table 4). This represents a slight, but not statistically significant increase (7.9%) over the number of persons who fished in the 12 months prior to this period. In terms of effort, Tasmanian residents accounted for over 64, fisher days of effort during the 12 month diary period. Overall 25.7% of fishers fished at least once in freshwater while 88.3% fished at least once in saltwater, with 23.2% of the effort (fisher days) involving freshwater fishing and 76.8% fishing in saltwater (Table 4). Table 4. Estimated number of persons and days fished by Tasmanian residents aged five years or older who fished in freshwater and saltwater in Tasmania during SE is standard error Freshwater Saltwater Total Effort Number SE Number SE Number SE Persons 32, ,884 3, ,781 4,79 Fisher days 149,248 13, ,655 24, ,489 28, Days fished In recreational fisheries, most fishers typically do relatively little fishing (and catch few fish) while, at the other extreme; relatively few fishers are very active and contribute disproportionately to the overall effort (and catch). The distribution of fishing effort is, therefore, characteristically highly skewed. Consistent with this general observation, about 7% of all fishers (some 92, persons) were estimated to have fished five or fewer days over the 12-month survey period while just 3% (about 3,4 persons) fished more than 2 days (Fig. 5). The median number of days fished per person was five for the survey period. TAFI Report Page 15